Agent-based Modelling to Evaluate the Impact of Plug-in Electric Vehicles on Distribution Systems

Abstract
Massive adoption of Electric Vehicles (EVs) could create issues in the electrical distribution system operation, in terms of currents and voltages. The analysis of the EV impact is a complex task, as the EV loads are variable in space and time depending on people routines, traffic conditions and recharge strategies. In this paper, an agent-based tool is presented to study the evolution of a city system with different EV penetrations. Drivers are represented through the use of several sociodemographic and psychological parameters in order to recreate a realistic activity schedule based on a set of rules. The system representation is divided into a static part referring to the environment, and a dynamic part where EV agents interact with each other and recharge their EVs influencing the grid. The results show the soundness of the approach and highlight where and when possible grid congestion can appear. Branch overloading, especially with high R/X ratio lines, is the first dangerous situation that can occur in distribution grids increasing the number of EV users.